HORTSCIENCE 40(7):2007–2010. 2005.
Transport of Cross-pollen by
Bumblebees in a Rabbiteye Blueberry
Planting
Patricio A. Brevis1
Department of Horticulture, Michigan State University, A316 Plant and Soil
Science Building, East Lansing, MI 48824-1325
D. Scott NeSmith2
Department of Horticulture, University of Georgia, Griffin Campus, Griffin,
GA 30223
Additional index words. Vaccinium ashei, Ericaceae, pollen dispersion, self-pollination,
Bombus spp., tetrad diameter
Abstract. Rabbiteye blueberry (Vaccinium ashei Reade) is a bee-pollinated small fruit crop
that often exhibits poor fruit set. Mixed cultivar plantings are recommended because crosspollination is required for optimum yields, and bees are expected to transfer pollen from
one cultivar to another. The objective of this study was to assess transport of cross-pollen
by bumblebees in a rabbiteye blueberry planting. Experiments were conducted in 2003 and
2004 in a plot composed of ‘Brightwell’ and ‘Climax’ plants arranged in alternating rows.
The proportion of ‘Brightwell’ and ‘Climax’ pollen carried on the bodies of bumblebees
was estimated based on frequency distributions of pollen diameter, measured with a particle
counter. About 75% of bumblebees collected in 2003 carried <20% cross-pollen. Proportions of cross-pollen in 2004 were higher than in 2003, but still, about 85% of bumblebees
collected carried <40% cross-pollen. The proportion of cross-pollen carried by bumblebees
changed during the flowering season. The greatest likelihood for cross-pollination occurred
during the time of maximum bloom overlap, although the median proportion of cross-pollen
was not >30% on any sampling day of 2004. The results from this study emphasize the need
to select more self-fertile rabbiteye blueberry cultivars and to maximize bloom overlap in
blueberry plantings.
The degree of self-fertility in many flowering plants is often limited by mechanisms such
as prezygotic self-incompatibility and earlyacting inbreeding depression (Klekowski,
1988; De Nettancourt, 1977; Seavey and Bawa,
1986). Outcrossing is thus beneficial or even
required in such species for fruit set and seed
maturation. In blueberries (Vaccinium section
Cyanococcus), cross-pollination normally
results in higher fruit set, larger fruit size, and
faster ripening rate relative to self-pollination
(El-Agamy et al., 1981; Meader and Darrow,
1944; Meader and Darrow, 1947; Morrow,
1943). Although highbush blueberries (Vaccinium corymbosum L.) can be planted in large
monoclonal blocks, yields obtained under these
conditions may not be as great as those from
mixed cultivar plantings (Hanson and Hancock,
1988). In rabbiteye blueberry (Vaccinium
ashei), cross-pollination is particularly critical
Received for publication 11 May 2005. Accepted for
publication 12 July 2005. The authors are greatly
appreciative of the financial support provided by the
Georgia Fruit and Vegetable Foundation. We also
thank Blair Sampson for providing assistance with
identification of bumblebee specimens and for his
critical review of the manuscript. A contribution of
the University of Georgia Agricultural Experiment
Stations, Georgia Station, Griffin. This research
was supported, in part, by state and Hatch Act funds
allocated to the Georgia Agricultural Experiment
Stations.
1
Research associate.
2
Professor.
HORTSCIENCE VOL. 40(7) DECEMBER 2005
for adequate fruit set (El-Agamy et al., 1981).
Therefore, interplanting two or more cultivars
is recommended to facilitate outcrossing and
improve yields.
Blueberries are pollinated by several species
of bees (Hymenoptera: Apiformes), and these
insects are expected to transfer pollen from one
cultivar to another. The information available
for blueberries, although scarce, suggests that
bee-assisted pollen dispersion between cultivars
may not be optimal. Vander Kloet and Lyrene
(1987) indicated that geitonogamous selfing
(pollen transfer between flowers in the same
plant) is likely to result as a consequence of
the foraging behavior of blueberry pollinators.
Moreover, Hancock et al. (1989) established that
fruit size and seed number per berry declined
with increasing distance from the source of
cross-pollen.
One of the most important horticultural
problems of the rabbiteye blueberry industry is
poor fruit set (Scherm et al., 2001). Although
knowledge of pollen dispersal is essential for
maximizing cross-pollination and achieving
optimal planting designs (Kron et al., 2001),
pollen dispersion between rabbiteye cultivars
has not been quantified previously. The objective of this research was to assess transport
of cross-pollen by bumblebees in a rabbiteye
blueberry planting.
Materials and methods
Study site. Experiments were conducted
in 2003 and 2004 in a mixed ‘Brightwell’ and
‘Climax’ rabbiteye blueberry plot located at
the Georgia Experiment Station in Griffin. The
plot was 33 × 29 m in size. Plants were spaced
at 3.7 m across rows and 1.5 m within rows.
‘Brightwell’ and ‘Climax’ plants were arranged
in alternating rows (i.e., 1:1 planting). The study
site was isolated from other sources of blueberry
pollen. The closest blueberry plantings were
>800 m away.
Proportion of cross-pollen carried by
bumblebees. Bumblebees (Bombus spp.) were
chosen as a model species to study pollinator
activity in blueberry plantings. The proportion
of cross-pollen carried by bumblebees was used
as an indirect measurement of pollen dispersion between ‘Brightwell’ and ‘Climax’ plants.
Cross-pollen is defined here as ‘Brightwell’
pollen transported by bumblebees to ‘Climax’
plants, and vice versa. Frequency distributions of pollen tetrad diameter, measured with
a particle counter, were used to predict the
proportion of ‘Brightwell’ and ‘Climax’ pollen
carried on the bodies of bumblebees. Brevis et
al. (2005) described the technique and statisti-
Table 1. Number of bumblebees collected per sampling day from ‘Brightwell’ and ‘Climax’ blueberry
plants in 2003 and 2004.
Specimens collected
per cultivar (no.)
Climax
Sampling date
Brightwell
Total
2003
21 Mar.
0
1
1
23 Mar.
0
8
8
29 Mar.
7
7
14
2 Apr.
10
5
15
14 Apr.
11
2z
13
Total
28
23
51
2004y
24 Mar.
1
7
8
27 Mar.
8
8
16
2 Apr.
23
12
35
5 Apr.
23
13
36
12 Apr.
31
0z
31
Total
86
40
126
z
Few or no specimens were collected on the last sampling day of each year due to the relatively low availability of open ‘Climax’ flowers.
y
Increasing numbers of bumblebees collected throughout the season reflect in part increasing pollinator
abundance.
2007
cal analysis used in this study, so only a brief
summary is given here. Tetrad diameter and
number were analyzed with a Coulter counter
(Multisizer II; Beckman Coulter, Fullerton,
Calif.). Cultivar proportions in pollen mixtures
carried by bumblebees were predicted using the
maximum likelihood method. Essentially, this
statistical analysis estimated the proportions of
‘Brightwell’ and ‘Climax’ pollen that would be
necessary to obtain the frequency distributions
of tetrad diameter observed in pollen mixtures
extracted from the bodies of bumblebees.
Sampling of bumblebees. Naturally occurring bumblebee queens were collected randomly
from ‘Brightwell’ and ‘Climax’ plants. Bombus
impatiens Cresson was the most abundant, but
B. griseocollis (Degeer), B. bimaculatus Cresson and B. nevadensis auricomus (Robertson)
were also collected. A detailed description of the
sampling procedure used in this study is given
by Brevis et al. (2005). The total number of
specimens collected in 2003 and 2004 was 51
and 126, respectively. Sample sizes per cultivar
and date are shown in Table 1.
Flowering phenology. ‘Climax’plants generally bloom earlier than ‘Brightwell’(Krewer and
NeSmith, 2000). To identify dates of cumulative
50% bloom and the period of maximum bloom
overlap, flowering phenology data were collected at the study site in both years. Three stems
were marked in each of 5 representative plants
of each cultivar. Open flowers were counted in
tagged stems at 1- to 3-d intervals during the
flowering season. Open flowers were removed
after recording data, so that individual flowers
could not be counted more than once. A nonlinear logistic model of the form y = N/[N + (1
– N) × exp(–r × t)] was fitted to the cumulative
bloom data using NLIN procedure of SAS (SAS
Institute, Cary, N.C.), where y was cumulative
bloom, t was time in days, and N and r were
parameters. The model gave a good fit to the
data (P < 0.0001 for each cultivar and year).
Fitted models were used to determine dates of
cumulative 50% bloom and to predict the daily
percentage of open flowers. Curves of open
flowers (≤5 d after anthesis) were simulated for
‘Brightwell’ and ‘Climax’ in 2004.
Fig. 1. Distributions of proportions of cross-pollen carried by bumblebees visiting ‘Brightwell’ and ‘Climax’ blueberry plants in a mixed rabbiteye blueberry plot in 2003 and 2004. Sample sizes (number of
bumblebees collected) appear above each bar.
Results
The proportion of cross-pollen carried
by bumblebees (mostly Bombus impatiens)
visiting ‘Brightwell’ and ‘Climax’ plants
in a mixed blueberry plot is summarized in
Fig. 1. In 2003, proportions of cross-pollen
from bumblebees collected in both cultivars
followed the same distribution (χ2 = 1.11; df
= 1, P = 0.29). About 75% of the specimens
collected carried <20% cross-pollen, and
bumblebees with a 50:50 pollen mix from each
cultivar were not found. In 2004, proportions
of cross-pollen for specimens collected from
‘Brightwell’ and ‘Climax’ did not follow the
same distribution (χ2 = 11.80; df = 2, P ≤
0.01). Most of the bumblebees collected from
‘Brightwell’ carried <20% cross-pollen, while
almost half of the catch from ‘Climax’ carried
between 20% and 40% cross-pollen.
In 2004, cumulative 50% bloom for ‘Brightwell’ occurred 5 d later than for ‘Climax’ (3
2008
Fig. 2. (A) Daily percentage of open flowers for each blueberry cultivar. The curves represent the predicted proportion of open flowers at stage ≤5 d after anthesis. (B) Median proportion of cross-pollen carried by bumblebees
visiting ‘Brightwell’ and ‘Climax’ plants in a mixed blueberry plot. Both graphs show data from Spring 2004.
HORTSCIENCE VOL. 40(7) DECEMBER 2005
Apr. and 29 Mar., respectively). The daily
percentage of open bloom and the median proportion of cross-pollen carried by bumblebees
are shown in Fig. 2. On 24 Mar., bumblebees
collected from ‘Climax’ carried on average
about 10% cross-pollen (that is, about 90%
‘Climax’ pollen). This was expected since
‘Climax’ was virtually the only cultivar with
open flowers on that day. The proportion of
cross-pollen carried by bumblebees foraging
on ‘Climax’ increased as more open ‘Brightwell’ flowers became available. The highest
likelihood for cross-pollination of ‘Climax’
flowers was recorded on 5 Apr. Specimens were
not collected later than 5 Apr. on this cultivar
because of the relatively low availability of
open bloom.
Bumblebees collected from ‘Brightwell’
plants on 27 Mar. carried, on average, only 5%
cross-pollen. This was intriguing since there
was a relatively high availability of ‘Climax’
flowers at the study site. The median proportion
of cross-pollen for pollinators in ‘Brightwell’
followed the same pattern observed for those
in ‘Climax’, and the maximum value was also
recorded on 5 Apr. On the last sampling day,
bumblebees visiting ‘Brightwell’ flowers carried about 10% cross-pollen. When sampling
was done simultaneously in both cultivars
(from 27 Mar. to 5 Apr.), bumblebees from
‘Brightwell’ always carried less cross-pollen
than those from ‘Climax’.
Bumblebees foraging on ‘Brightwell’ carried more blueberry tetrads on their bodies than
those on ‘Climax’. In 2004, the mean number
of tetrads (± SE) extracted per bumblebee was
917 ± 93 and 539 ± 49 for specimens collected
from ‘Brightwell’ and ‘Climax’, respectively
(significantly different at P < 0.01). These
results confirmed the trend observed in 2003.
The total number of tetrads and the proportion
of ‘Brightwell’ pollen carried by bumblebees
in 2004 were weakly but positively associated (Fig. 3). The association between the
log-transformed number of tetrads and the
proportion of ‘Brightwell’ was linear (r = 0.24,
P ≤ 0.01; n = 126).
Discussion
According to Shutak and Marucci (1966),
fruit set levels as high as 80% are required for
adequate commercial production of highbush
blueberry. In rabbiteye blueberry, estimates as
low as 8% to 10% have been recorded for the
widely planted cultivar Tifblue (Lyrene and
Crocker, 1983; NeSmith and Adair, 2004).
Since fruit set responds positively to increased
bee density (Dedej and Delaplane, 2003), it
is generally thought that fruit set of rabbiteye
blueberry is limited by insufficient pollen
delivery to the stigmas. The results from the
present study indicate that, in addition to insufficient pollen quantity, the source of the pollen
deposited by pollinators could also play a role
in determining fruit set under field conditions.
Pollinators are likely to carry high proportions
of self-pollen, and blueberry pistils do not have
mechanisms to reject it (El-Agamy et al., 1982;
Garvey and Lyrene, 1987; Krebs and Hancock,
1988; Vander Kloet, 1991). In cultivars with a
low degree of self-fertility, deposition of pollen
Fig. 3. Total number of tetrads and proportion of ‘Brightwell’ pollen carried by bumblebees visiting
‘Brightwell’ and ‘Climax’ blueberry plants in a mixed blueberry plot during Spring 2004. The line was
fitted across untransformed data. Coefficient of correlation and significance level were estimated using the
log-transformed number of tetrads per bumblebee.
HORTSCIENCE VOL. 40(7) DECEMBER 2005
loads that are rich in self-pollen would represent a major limitation for fruit set. However,
cultivars with a higher degree of self-fertility
could still exhibit adequate fruit set under
similar circumstances, even though fruit size
and ripening rate may not be optimum. These
findings indicate that breeding efforts should
be made to select more self-fertile rabbiteye
blueberry cultivars. Also, additional research
is needed 1) to establish the effect of planting
arrangements on pollen dispersion and 2) to
determine if certain bee species carry more
cross-pollen than others.
Bumblebees with a 50:50 pollen mix from
each cultivar were rare in 2004 and absent
in 2003. These results clearly indicate that
bumblebees visited ‘Brightwell’ and ‘Climax’ flowers in a nonrandom fashion. Waser
and Price (1983) reported that 90% or more
of bumblebees’ flower-to-flower flights are
shorter than 1 m. Low proportions of crosspollen carried by bumblebees at the study site
might be explained by their tendency to visit
many flowers within a plant (Vander Kloet
and Lyrene, 1987) and to fly between adjacent
monoclonal plants within the same row.
Proportions of cross-pollen carried by
bumblebees changed during the flowering
season. The greatest likelihood for cross-pollination occurred around the date of maximum
bloom overlap. Therefore, the chances of
cross-pollination at the study site were not
only low, but were also limited to a narrow
time window. The results from this study emphasize the need to maximize bloom overlap
in blueberry plantings.
The low proportion of cross-pollen carried
by bumblebees collected from ‘Brightwell’
early in the season was not consistent with the
availability of open ‘Climax’ flowers. Selectivity toward virgin ‘Brightwell’ flowers could
explain these results, although no data were
collected to determine cultivar preferences
by bumblebees.
Brevis and NeSmith (2004) established that
‘Brightwell’, on a per flower basis, releases
more tetrads than ‘Climax’. This difference
seems to have an effect on bumblebee-assisted pollen carryover, and thus, on the extent
of cross-pollination. Data from 2004 indicate
that bumblebees harvested more tetrads from
‘Brightwell’ than from ‘Climax’. Bumblebee
foraging behavior is such that when they assess
high pollen returns during a flower visit, both the
frequency of grooming episodes and the handling duration per flower increase (Buchmann
and Cane, 1989; Harder, 1990). Pollen carryover
by pollinators during subsequent flower visits
is thus restricted, because pollen is packed into
scopae instead of being transported to the next
flower. In concordance with this, bumblebees
visiting ‘Brightwell’ flowers in 2004 carried
lower proportions of cross-pollen relative to
those collected from ‘Climax’. The results from
this study suggest that blueberry cultivars that
release more pollen per flower can actually be
exposed to a higher degree of selfing.
Literature Cited
Brevis, P.A. and D.S. NeSmith. 2004. Pollination
discoveries on rabbiteye blueberries, p. 125–131.
2009
Proc. Ga. Fruit and Veg. Growers Conf. 9–10
Jan., Savannah.
Brevis, P.A., D.S. NeSmith, L. Seymour, and D.B.
Hausman. 2005. A novel method to quantify transport of self- and cross-pollen by bees in blueberry
plantings. HortScience 40(7)2002–2006.
Buchmann, S.L. and J.H. Cane. 1989. Bees assess
pollen returns while sonicating Solanum flowers.
Oecologia 81:289–294.
Dedej, S. and K.S. Delaplane. 2003. Honey bee
(Hymenoptera: Apidae) pollination of rabbiteye
blueberry Vaccinium ashei var. ‘Climax’ is pollinator density-dependent. J. Econ. Entomol.
96:1215–1220.
El-Agamy, S.Z.A., W.B. Sherman, and P.M. Lyrene.
1981. Fruit set and seed number from self- and
cross-pollinated highbush (4x) and rabbiteye
(6x) blueberries. J. Amer. Soc. Hort. Sci.
106:443–445.
El-Agamy, S.Z.A., W.B. Sherman, and P.M. Lyrene.
1982. Pollen incompatibility in blueberries (Vaccinium spp.). J. Palynol. 18:103–112.
Garvey, E.J. and P.M. Lyrene. 1987. Self incompatibility in 19 native blueberry selections. J. Amer.
Soc. Hort. Sci. 112:856–858.
Hancock, J.F, S.L. Krebs, M. Sakin, and T.P. Holtsford.
1989. Increasing blueberry yields through mixed
variety plantings. Mich. State Hort. Soc. Annu.
Rpt. 119:130–133.
Hanson, E.J., and J.F. Hancock. 1988. Hints on
2010
growing blueberries. Mich. State Univ. Ext.
Bul. E-2066.
Harder, L.D. 1990. Behavioral responses by bumble
bees to variation in pollen availability. Oecologia
85:41–47.
Klekowski, Jr., E.J. 1988. Mutation, development,
selection, and plant evolution. Columbia Univ.
Press, New York.
Krebs, S.L. and J.F. Hancock. 1988. The consequences
of inbreeding on fertility in Vaccinium corymbosum L. J. Amer. Soc. Hort. Sci.113:914–918.
Krewer, G. and D.S. NeSmith. 2000. Blueberry cultivars for Georgia. Univ. Ga. Fruit Publ. 00-2. 25
Mar. 2005. http://www.smallfruits.org/Blueberries/production/BlueberryCultiv.htm.
Kron, P., B.C. Husband, P.G. Kevan, and S. Belaoussoff. 2001. Factors affecting pollen dispersal
in high-density apple orchards. HortScience
36:1039–1046.
Lyrene, P.M. and T.E. Crocker. 1983. Poor fruit set
on rabbiteye blueberries after mild winters: Possible causes and remedies. Proc. Fla. State Hort.
Soc. 96:195–197.
Meader, E.M. and G.M. Darrow. 1944. Pollination of
the rabbiteye blueberry and related species. Proc.
Amer. Soc. Hort. Sci. 45:267–274.
Meader, E.M. and G.M. Darrow. 1947. Highbush
blueberry pollination experiments. Proc. Amer.
Soc. Hort. Sci. 49:196–204.
Morrow, E.B. 1943. Some effects of cross-pollination
versus self-pollination in the cultivated blueberry.
Proc. Amer. Soc. Hort. Sci. 42:469–472.
NeSmith, D.S. and H.M. Adair. 2004. Rabbiteye
blueberry field trials with the growth regulator
CPPU. Small Fruits Rev. 3(1/2):183–191.
De Nettancourt, D. 1977. Incompatibility in angiosperms. Springer, Berlin.
Scherm, H., D.S. NeSmith, D.L. Horton, and G.
Krewer. 2001. A survey of horticultural and pest
management practices of the Georgia blueberry
industry. Small Fruits Rev. 1:17–28.
Seavey, S.R. and K.S. Bawa. 1986. Late-acting
self-incompatibility in angiosperms. Bot. Rev.
52:195–219.
Shutak, V.G. and P.E. Marucci. 1966. Plant and fruit
development, p. 179–198. In: P. Eck and N.F.
Childers (eds.). Blueberry culture. Rutgers Univ.
Press, New Brunswick, N.J.
Vander Kloet, S.P. 1991. The consequences of mixed
pollination on seed set in Vaccinium corymbosum.
Can. J. Bot. 69:2448–2454.
Vander Kloet, S.P. and P.M. Lyrene. 1987. Self-incompatibility in diploid, tetraploid, and hexaploid Vaccinium corymbosum. Can. J. Bot. 65:660–665.
Waser, N.M. and M.V. Price. 1983. Optimal and
actual outcrossing in plants, and the nature of
plant-pollinator interaction, p. 341–359. In:
C.E. Jones and R.J. Little (eds.). Handbook of
experimental pollination biology. Van Nostrand
Reinhold, New York.
HORTSCIENCE VOL. 40(7) DECEMBER 2005